Abstract
In the theory of the Behavioral Inhibition System (BIS), Gray and McNaughton (2000) classified events that produce or inhibit goal-directed behaviour into two affective categories: approach versus avoidance. We expenence goal-conflict when approximately equal but incompatible approach and avoidance tendencies are concurrently activated. Gray and McNaughton (2000) proposed goal-conflict as a class of mechanisms separable from "simple" mechanisms: Goal-conflict effects are maximal when incompatible approach and avoidance tendencies are balanced. Simple effects are maximal when either approach or avoidance tendencies predominate.
Gray and McNaughton (2000) saw the hippocampus as a key nexus for resolving goal-conflict by recursive amplification of the subjective value of punishment, thereby increasing avoidance tendencies. Rodent hippocarnpal theta (4-12 Hz) is necessary (but not sufficient) for correct and efficient transmission of hipppocampal outputs. The BIS theory is fundamentally an animal model. It is not clear if a human BIS exists in the same form. Record human hippocampal (4-12 Hz) activity from the scalp is unlikely. However, during goal-conflict resolution, cortically generated theta recorded from the scalp could be modulated by human hippocampal theta. Therefore, superficially recorded 4-12 Hz theta spectra power was used to assess if specific goal-conflict processing activity could be detected in humans.
Human goal-conflict processing was assessed in four experiments: the Stop-Signal Task (SST), an existing experimental task, and three variations of a task termed "Choice", created for this thesis. Across experiments, three key conditions were created. Approach and avoidance were balanced in the intermediate condition (maximal goal-conflict). Net approach and avoidance predominated in the adjacent conditions respectively (minimal goal-conflict). Goal-conflict was assessed as the difference between activity in the intermediate condition and the average activity across the adjacent conditions (via extraction of the orthogonal quadratic trend for significance testing).
Goal-conflict increased activations consistently at F8, above the right frontal cortex. Increase in task dependent goal-conflict activations were also observed at F7, Fz and F4 above the frontal cortex, and T3, T4, T5 and T6 above the temporal cortex. Activations within the human theta frequency range (4-7 Hz) were consistently detected in the Choice tasks. In the SST, activations spanned the conventional human theta (4-7 Hz) and alpha (8-12 Hz) frequencies. In the Choice tasks, higher conflict theta at T3, T5 and F8 predicted increased avoidance.
Taken together, the findings support Gray and McNaugthon's (2000) views that a) goal-conflict is a class of mechanism separable from simple approach and avoidance; b) goal-conflict processing recruits and increases cortical rhythmic activity within the same frequency range as rodent hippocampal theta (4-12 Hz); and goal-conflict is resolved by increasing the subjective value of punishment, thereby increasing avoidance tendencies. Although speculative, the current work identified a right inferior frontal gyrus neural circuit for slower, and a presupplementary motor area circuit for faster behavioral inhibition during goal-conflict resolution. These circuits are not explicit in the current BIS model.